In her lecture, “Kant on Reason and Religion” (Delivered at Harvard University, April 1—3, 1996), Onora O’Neill gives the following account of Kant’s requirements for discipline of reason:
Kant’s account of the discipline of reason can be summarized in three claims. First, in calling reason a discipline, he is claiming that it is a negative constraint on the ways in which we think and act: there are no substantive axioms of reason, whose content can fully steer processes of reasoning; there are merely constraints. Reason is indeed merely formal.
Second, the discipline of reason is nonderivative. Reason does not derive from any more fundamental standards. On the contrary, it appeals to no other premises, so can be turned on any claim or belief or proposal for action. Neither church nor state, nor other powers, can claim exemption from the scrutiny of reason for their pronouncements and assumptions. The authority of reason would be nullified by any supposition that it is subordinate to the claims of one or another happenstantial power…
If reason has any authority, it must be its own rather than derivative.
Although reason does not have derivative authority, authority it must have. Authority is needed to distinguish between ways of organizing thought and action that are to count as reasoned and those that are to be dismissed as unreasoned. Kant traces this nonderivative authority to the requirement that reasons be public, in the sense that they can be given or exchanged, shared or challenged. Nothing then can count as reasoned unless it is followable by others, that is, unless it is lawlike. Ways of organizing thought and action that are not lawlike will be unfollowable by at least some others, who will view them as arbitrary or incomprehensible.
The minimal, modal requirement that reasons be followable by others, without being derivative from other standards, is Kant’s entire account of the authority of reason. Yet mere nonderivative lawlikeness has considerable implications for the organization of thought and action: in the domain of theory it amounts to the demand that reasons be intelligible to others; in the domain of action it amounts to the requirement that reasons for action be ones that others too could follow.O’Neill also points out that “the supreme principle of practical reason is presented as a negative (formal) requirement that is underivative because it appeals to no other spurious “authorities” (that would be heteronomy) and demands adherence to lawlike maxims (i.e., to maxims that could be adopted by all).”